The Joint Automated Booking System
Audit Report 05-22
Office of the Inspector General
The goals and mission of the JABS initiative varied somewhat from the initial effort under the Pilot Project to the current implementation. For the Pilot Project, the JABS goals were to: 1) streamline the booking process through automation and elimination of duplication, 2) allow updates to prisoner data, 3) standardize data collected, and 4) improve the process to identify repeat offenders and persons with outstanding charges. The Department considered the Pilot Project a success and decided JABS should be implemented nationwide.
The Executive Summary of the 1999 Boundary Document described the major goals of JABS as being to: 1) automate the booking process, 2) enable each law enforcement organization to share and exchange booking information, and 3) establish a federal offender tracking system. These goals were jointly accepted by officials from JMD, the BOP, FBI, DEA, USMS, and INS as the basis of establishing requirements for the nationwide JABS, guiding the development and implementation of JABS, and expressing the commitment of and agreement between the JABS Board of Directors and the JABS Program Management Office.
The JABS Program Management Office also projected in the Boundary Document that JABS would ensure the rapid positive identification of offenders, reduce or eliminate redundant booking steps, and aid in criminal investigations. We consider the rapid positive identification of offenders to be a component of Goal 1 and the intentions to eliminate redundant booking steps and aid in criminal investigations to be components of Goal 2. The discussion below summarizes the progress made in accomplishing each of these goals.
The JABS Boundary Document describes goals and objectives related to automating the collection of offender data, providing connectivity to each component's ABS, and ensuring the rapid positive identification of offenders through the FBI's IAFIS. The JABS program projected that achieving these goals would result in decreased fingerprint transmittal time to the IAFIS, decreased rejection rates for fingerprint sets submitted to the FBI, and a streamlined booking process.
To automate this process, the JABS program needs to deliver the capability for each Department law enforcement component to: 1) electronically record standardized booking data, including digitized fingerprints and photographs; 2) transmit usable data to IAFIS securely for identification; and 3) receive responses from IAFIS quickly. Streamlining the process within each component could result from reusing information that has been entered once into the ABS.
We found that the JABS program has automated the booking process in each Department law enforcement component and the DHS, and provides the necessary connectivity to send fingerprints to the FBI for identification, and receive responses quickly from the FBI. As we discuss in Finding 2, these capabilities are available at many, but not all, of the Department's booking locations. We also found that JABS recently added the capability for the USMS to submit bookings to JABS and fingerprints to IAFIS on behalf of organizations and locations that do not have access to their own ABS.
In terms of outcomes, we found JABS has resulted in significantly decreased transmittal time for fingerprints to the FBI, which, in turn, makes a response available within hours rather than weeks. Fingerprint sets sent to the FBI through JABS also have a slightly lower rejection rate than all criminal hard-copy sets submitted, and additional improvements may be possible from future technology advances. The booking process streamlining has been partially achieved by reducing the number of times a component has to take a set of fingerprints. However, participating agencies still manually complete hard-copy forms, containing information that has already been recorded in the automated booking station, which are then brought to the USMS detention facility when the offender is transferred to USMS custody. The details supporting these conclusions follow.
All of the component automated booking systems run on workstations consisting of desktop computers attached to cameras, digital fingerprint-capture devices (live scanners), flat-bed scanners (for scanning hard-copy fingerprint cards), and printers. The workstations provide a graphical user interface from which users select various booking tasks to perform, such as capturing fingerprints and photographs. Once recorded in the ABS, fingerprints and photographs can be printed out as many times as the user chooses.
Those agencies with an interface from their legacy system automatically populate data elements in the ABS screens with the appropriate data when a user begins work on that record. The booking agents then record data elements that have not come over from another system, take the offender's fingerprints and capture photographs. Before transmitting the data to Core JABS to be validated and forwarded to IAFIS, users check that the package meets certain standards through validation routines on the booking station. They can make corrections or changes to the information, or re-take fingerprints, if necessary. The fingerprint quality screening identifies specific fingers that do not pass the quality tests, and only the fingers at issue need to be re-rolled. When the ABS record is complete and has been validated, it is transmitted to the Core JABS. Users recording information in an ABS are not connected to the central JABS database.
Through the ABS, users can process different types of transactions that vary somewhat between participating agencies. The transaction types are described below.10
Bookings - Bookings are used to record a federal arrest, submit criminal information to the FBI's IAFIS and JABS, and obtain identification and a criminal history (commonly called a rap sheet) from the FBI. The FBI's identification is based on 10 rolled fingerprints and 4 flat images. A positive identification from IAFIS requires an FBI fingerprint specialist to verify any match that the system has identified. A booking transaction receives an identification response from IAFIS, and the information is retained in JABS and IAFIS.
Second-tier bookings are performed by USMS detention facilities when arresting agencies transfer custody of prisoners to the USMS. These bookings are identified by JABS as "criminal history" transactions, but the USMS refers to them as bookings. This transaction type generates the same response from IAFIS as a booking, so the USMS receives the identification response and rap sheet. The difference is that it is not stored in IAFIS. It is stored in JABS, along with data input by the USMS. The same applies to the third-tier bookings performed by the BOP to obtain positive identification from IAFIS upon the initial commitment of an offender to a BOP facility.
Interagency Bookings - This transaction type, implemented in June 2004, is designed to allow the USMS to submit booking data to IAFIS on behalf of other arresting agencies. Any arrest made by a federal law enforcement agency that routinely brings suspects to the USMS for booking can now be entered into JABS, and therefore IAFIS, through a USMS detention facility that has been equipped with this version of the ABS software. The identification response and the rap sheet from IAFIS can be transmitted to both the arresting officer and the USMS location. The arresting agency no longer needs its own access to JABS.
Updates - Arresting agencies and the USMS may submit additional information or changes to records that have been created in JABS through bookings and criminal history transactions. JABS verifies that an update transaction is associated with an original transaction by comparing the fingerprint images for the package identifier.
Inquiries - An inquiry is used to obtain a criminal history on a person, based on an FBI fingerprint identification, either prior to an arrest or in other law enforcement activities not associated with an arrest, for example, to obtain information about confidential informants. The information is not stored in either JABS or IAFIS. The DHS calls this a "search with verification."
Search, Quick Search, or Ten-Print Rap Sheet Requests (TPRS) - This transaction, which the DHS calls a "quick search," is based on a service the FBI performs only for the DHS to help meet the specialized needs of immigration and border officials who need to quickly identify any reasons for denying entry to persons trying to enter the U.S. The search is based on 10 rolled prints and minimal biographical data. It is a completely automated search of the IAFIS fingerprint files, without human verification. It returns possible matches, rather than a verified identification, to the DHS very quickly. According to the FBI, it is rare for IAFIS to present more than one possible match. After the DHS receives the information, officers can determine whether to go ahead with a search with verification, and then a booking, depending on the circumstances. According to DHS officials, this transaction type currently supports part of the IDENT/IAFIS project by allowing DHS to search IAFIS and receive a fast response.
As of November 2004, the following participating agency booking stations had been brought online.
The ATF Automated Booking Station (ATF ABS) was linked to the Core JABS in December 2003, and the ATF began submitting bookings from one site in January 2004.11 As of November 2004, the ATF was still in the process of deploying the majority of its planned booking stations. This first booking station, located in Richmond, Virginia, was part of a pilot test and was based on the ABS in use at FBI field offices. The ATF told us its ABS receives some data elements from the ATF's legacy database system, N-FORCE, needed to populate JABS data elements. Users enter supplemental information into the ABS, including fingerprints and photographs. According to the JABS Program Management Office, the ATF is currently deploying the first agency-wide ABS to use the new web-based version of JABS.
The BOP Automated Booking Station (BOP ABS) is linked to the BOP's legacy data system, SENTRY. SENTRY is an online, real-time, database management system that monitors the system-wide movement and management of inmates, including sentence computations, work assignments, institution designation, administrative remedies, discipline, and inmate financial responsibility. The BOP's automated booking station functions as an integrated part of the SENTRY system, and was first linked to the Core JABS in October 2002, although successful operations did not begin until April 2004.
The DEA Firebird Booking Service (FBS) is part of DEA's Firebird information system infrastructure. Firebird includes a Case Management System that provides a custom user interface for creating, editing, storing, retrieving and printing case-related material. The DEA developed its automated booking station as a component of the Firebird Case Management System and linked it to the Core JABS in July 2000. The DEA is the only component that uses a version it calls "JABS Lite" at smaller offices. The difference from a standard ABS workstation is that the Lite version comes with only a flatbed scanner to scan in hard-copy fingerprint cards. There is no live-scan fingerprint device attached to the station.
The FBI Automated Booking Station (FBI JABS) is a stand-alone work station that is not connected to any other FBI system or network. It transmits data using dial-up connections to the Core JABS. This ABS was based on the DEA's Firebird Automated Booking Station to leverage software that had already been developed, and was modified to meet the FBI's requirements. The FBI linked the FBI ABS to the Core JABS in August 2002.
The USMS Automated Booking Station (USMS ABS) is interfaced with the USMS legacy data system, the Prisoner Tracking System (PTS), which is an old application running on a network of distributed application and database servers. The USMS officers enter information about an offender first into the PTS. Some of that information is then migrated into the ABS when the user selects the ABS menu option during the booking procedure. The USMS linked its ABS to the Core JABS in August 2002.
The DHS IDENT/IAFIS captures data on persons being detained by the DHS for immigration purposes, and meets needs that vary significantly from the other law enforcement agencies.12 DHS users in the Bureau of Customs and Border Protection need to identify people accurately to decide whether to let them into the United States, or detain or arrest them. The DHS's primary legacy systems supporting border functions are the ENFORCE and IDENT systems. Many locations now use an integrated workstation in which users record data elements once and can access functions of the two legacy systems and IAFIS (through JABS). When the information has been recorded, users select the types of transactions to perform, which include three transaction types involving the Core JABS and IAFIS: the booking, quick search, and search with verification (or inquiry) described above. The initial version of this system was deployed first in September 2002.
After using the ABS, many arresting agencies still manually complete various hard-copy forms to accompany offenders when custody is transferred to the USMS. Different USMS facilities require variations of the forms. These forms contain data elements that have already been entered into the ABS. This duplication is further described in the section of this finding headed Sharing and Exchanging Booking Information.
The initial version of the Core JABS became operational in July 2000. In subsequent months, modifications to the Core JABS were placed into production incrementally, and in June 2004, Version 3.0, which offers a three-tier, web-based architecture, became available for future deployment.
The Core JABS is described in JABS system documents as a "store and forward" system that receives information from law enforcement agencies, stores some of the information in a database, and forwards the information to the FBI's IAFIS. Updates to booking or criminal history transactions coming into the Core JABS are not forwarded to IAFIS, but are maintained as separate transactions in the nationwide JABS database. After receiving a response from IAFIS, the Core JABS forwards the response to the original submitter of the information. JABS does not compare fingerprints to identify people.13
The architecture for the Core JABS is based on commercial off-the-shelf components and resides at the Justice Data Center outside Washington, D.C. The data center is operated by the Justice Management Division and provides planning, management, facility services, and operational support for the Department of Justice. Warm-site redundancy for JABS is provided at the Justice Data Center in Dallas, Texas.14
Interface with the FBI's IAFIS
JABS implementation has automated the transmittal of fingerprints to the FBI and the receipt of identification and criminal history responses from the FBI for locations at which an ABS has been deployed. Automating the collection and transmission of offender fingerprints was projected to decrease transmittal time to the FBI by 90 percent and decrease the percentage of fingerprints rejected by the FBI. We determined that the timeframes for submitting and receiving fingerprint information had decreased dramatically, and that the rejection rates for fingerprints by the FBI may have improved somewhat.
According to representatives of participating agencies we interviewed, fingerprint transmittal time to IAFIS has improved exponentially. The improvement is based on the nearly instantaneous transmittal of packages to the Core JABS, which, in turn, forwards each package to IAFIS. After receiving the response from IAFIS, the Core JABS returns the IAFIS response to the source within two minutes.
We determined that the time it takes from submitting fingerprints to the FBI and receiving a response has decreased from several weeks to a matter of hours, and sometimes minutes. Sources including the FBI and JABS Board representatives confirmed that before JABS the process took several weeks to receive a response back from the FBI. In many cases, that response may only inform the arresting officer that the fingerprints were not of acceptable quality to use for identification. Agency representatives and users agreed that they receive responses back very fast now, although the exact timeframes varied between agencies. Staff we interviewed indicated they were receiving responses generally within one to three or four hours, with the exception of DEA staff, who stated that the process may take as long as overnight. Staff at the JABS Program Management Office told us that this was likely a function of the DEA network. In any event, the improvement over mailed hard-copy fingerprints remains significant.
Additionally, DHS representatives told us that the response time from the FBI for quick search/TPRS transactions occurred within minutes, and could be as little as several seconds.
Fingerprint rejections by the FBI were projected to decrease with implementation of JABS. This expectation was based on administrative reasons for rejection, such as inconsistent data elements. However, according to the FBI, administrative errors cause less than 7 percent of the rejections. The use of software to assess the quality of fingerprint images at the point of entry and to interact with the user to obtain a good set of fingerprints should also be expected to reduce the number of rejections by the FBI.
In order for fingerprints to be usable by IAFIS, they must meet quality standards which conform to standards set by the National Institute of Standards and Technology and the FBI's Electronic Fingerprint Transmission Specification. Before attempting to match fingerprints through IAFIS, the FBI screens them for usability and rejects sets that do not meet the quality criteria.
Participating agencies' automated booking systems all use commercially available software to assess and validate the quality of fingerprint images. During the validation process, the software alerts the user to fingerprints that do not meet quality standards, and the user can re-record the identified fingerprints. These products have been approved by the FBI for the submission of fingerprints, but do not guarantee that 100 percent of submissions will be determined usable for IAFIS. Also, transmission anomalies might garble a small percentage of fingerprints during transmission.
The FBI, however, has not shared with the JABS program the quality screening process used in IAFIS, so each component's ABS does not pre-screen fingerprints with precisely the same quality validation process the FBI uses. The FBI told us that the National Institute of Standards and Technology is in the process of developing a non-proprietary standard for quality control of fingerprints, and that the FBI and vendors will move toward this standard once it is released. FBI officials indicated they prefer for the JABS program to wait for the non-proprietary standard to be released before implementing additional fingerprint quality control.
The ABS also allows users to override the fingerprint quality screening for individual prints, and submit a set of prints in which not all prints have passed the acceptance process. Booking officers told us that overriding the ABS process is sometimes necessary because of problems with a subject's fingers, such as scars and stains. During site visits and interviews, many users complained about the sensitivity of the live-scan equipment, and most agreed that it was difficult to get JABS to accept fingerprints from people with thin or short fingers. We discussed this concern with the JABS Program Management Office staff, who told us they are anticipating purchasing new generation live-scan equipment that will hopefully alleviate the problem.
Fingerprints may be accepted for processing by IAFIS, even when one or more of the prints have not passed the ABS screening, and FBI officials indicated the FBI has the ability to improve the usability of some images. The imperfect fit in procedures between the ABS and the FBI in screening the quality of fingerprint images results in the FBI continuing to reject some fingerprint submissions received from JABS. The fact that the ABS rejects fingerprints that may be accepted by the FBI may encourage users to override the ABS quality screening.
Although rejections of federal fingerprint submissions to IAFIS may have decreased since JABS implementation, we were not able to obtain baseline data about past performance by federal law enforcement on hard-copy submissions to measure any direct improvement. However, we were able to obtain data showing recent differences between electronic submissions from JABS, and all hard-copy criminal fingerprint submissions to IAFIS.
We obtained information from the JABS program and the FBI about fingerprint rejection rates for both JABS electronic transactions and hard-copy submissions. The FBI provided summary information that 3.4 percent of criminal fingerprint sets submitted electronically through JABS from October 1, 2003, through April 30, 2004, were rejected, compared with a rejection rate of 3.8 percent for all criminal hard-copy fingerprint submissions to IAFIS.15 This represents a difference of about 11 percent, favoring the JABS electronic submissions.16 This is not a direct measure of any improvement that may be experienced by federal law enforcement, but suggests that some improvement may result from automating the process.
We also reviewed summary data in the Core JABS Metrics Summary Reports for FYs 2002, 2003, and 2004 through July. IAFIS rejection rates reported by the Core JABS for FYs 2002, 2003, and 2004 through July are shown in the table below.17
IAFIS Rejection Rates
Although these numbers do not compare hard-copy and electronic submissions, they do clearly show a pattern of improvement over time in the electronic submissions themselves. It is possible that rejections may decrease further in the future as JABS technology improves and users become more practiced with the equipment.
FBI officials indicated that an additional quality checking procedure, sequence checking, might help to decrease the number of rejections even further. Sequence checking is a process that ensures fingerprints are in the proper location in the record by comparing flat four-finger images with individual finger images. This procedure was not included in the ABS quality screening of fingerprints during our audit.
An FBI CJIS representative told us that about 4 percent of rejections were due to sequence inconsistencies, not just from JABS, but from all submitters. The Program Management Office indicated it had begun implementing sequence checking around September 2004. The Acting Program Manager agreed with the FBI that a sequence validation check should further reduce the number of rejections from IAFIS.
For this audit, we assessed system performance in terms of the Core JABS availability and internal processing time. To determine if the Core JABS availability and internal processing times were adequate, we obtained criteria specific to both and compared the criteria to JABS internal reports detailing actual availability and processing times. In addition, we interviewed participating agency representatives and users about whether they had any problems with the availability or response times from JABS.
System Availability - The JABS program's performance goals and measures include making Core JABS available 24 hours a day, 7 days a week, because bookings can occur at all hours. We obtained summary information from the JABS Program Management Office for October 2002 through February 2004. The reported percentage of availability, by month, was between 91 and 100 percent for the entire period, with only two months of the entire period below 97 percent, as shown in the following graph. As of December 2004, the Program Management Office reported overall system availability of 99.5 percent for FY 2004, and 99.8 percent for the first two months of FY 2005.
Processing Time - We assessed internal processing time for the Core JABS, rather than overall response times to users in the field. The Core JABS internal processing time does not include the processing or transmission time outside of the firewall to the Core JABS, such as time within a participating agency's own network. The performance standard set for the system to forward packages to IAFIS upon receipt at the initial firewall, and to return an IAFIS response to the submitting agency's network is 120 seconds.
We obtained the Core JABS Processing Time reports for October 2001 through February 2004 to see if Core JABS internal processing times met the standard. The reports show a summary figure in seconds by month. We found that the internal processing times were generally less than 120 seconds before September 2003, when the processing times increased to between 130 and 200 seconds. The time came back down to less than 120 seconds in February 2004, as shown in the following graph.
JABS Processing Times
The JABS program manager told us the reduction in February was the result of a system improvement that was implemented to handle the exploding number of transactions being submitted by the growing number of users. Staff at the Program Management Office told us the final FY 2004 average processing time was 87 seconds, which was further reduced to 6 seconds for the first two months of FY 2005.
We interviewed representatives on the JABS Board of Directors from the BOP, ATF, DEA, DHS, USMS, and FBI to obtain their comments on how well JABS was working within their organizations. The representatives agreed that the most significant benefit of using JABS is the quick positive identification response from IAFIS.
To determine if the program's users were satisfied with the booking capabilities of JABS, we contacted staff from the major Department components and DHS staff who used the system on a regular basis. We did not contact BOP personnel because the BOP ABS was not operational for most of the period under review and staff had limited experience with the system. Department staff commented positively on the quick turnaround response time from IAFIS. The DHS staff was likewise positive about the ability to electronically book an offender and receive fast responses from IAFIS.
Because of the DHS's disproportionately high level of participation in JABS, we judgmentally selected 50 individuals from the universe of users to survey. We successfully contacted 45 users and asked how JABS and or the ability to access IAFIS met their expectations. We also asked the same individuals to provide a numerical rating (1-5, with 1 as poor) on the ability to electronically access and use the booking station to access IAFIS. An overwhelming majority (36 of 45, or 80 percent) rated the system either a 4 or a 5. The few who rated it lower stated they felt the electronic access to IAFIS was a great benefit, but that their booking station was down too often, which created hardships on their offices. Thirty-two of 45 individuals (71 percent) contacted stated the number of arrests at their facilities had increased because of their ability to access IAFIS.
As envisioned in the JABS Pilot Project and the JABS Boundary Document, JABS was intended to: 1) support booking automation through data sharing, thereby reducing or eliminating redundant booking steps, data, and data entry; and 2) allow components to share one another's offender data to aid in criminal investigations. The success of these efforts would be measured by a decrease of 50 percent in the time it takes to perform a subsequent booking, which is a second or third-tier booking by the USMS and BOP, and in decreased time to identify existing federal booking information.
To achieve these goals and objectives, the program would need to make previously recorded information available to multiple agencies to use for bookings and inquiries. To reduce or eliminate booking steps and data, booking information could be input by the arresting agency once, after which the second or third-tier agencies would access the common booking data and append or update the record with additional information as needed, including the disposition of the offender. To aid in criminal investigations, the automated information would be made accessible to users who are involved in investigations.
We found that JABS has not reduced or eliminated redundant booking steps, data entry, or data between participating agencies. Nor has the cycle time for subsequent bookings been decreased through data sharing, as envisioned in the Pilot Project and the Boundary Document. JABS has the ability to transfer the content of existing booking packages that are contained within JABS to any booking station. The USMS and BOP use their own case management systems, the PTS and SENTRY, to begin processing offenders. They re-enter in those systems information that may already exist in the Core JABS database because they cannot import data from the ABS into the PTS and SENTRY. The Query Tool was envisioned to export records to be used by each ABS. Records would be imported from the nationwide repository into the ABS to create new bookings. We found that no participating agency was importing data from the central JABS database to create new records.
We also found that the JABS program has made it possible for components to share data for investigative and research purposes through the JABS Query Tool. Investigators have online access to search the central JABS database, which makes information available immediately on offenders included in the database, thereby reducing the time it takes investigators to locate existing information. However, two important functions that would be useful to investigators were not in place during our audit. Fingerprint data was not available for viewing or printing, and the planned electronic photo line-up capability had not been implemented. The details supporting these conclusions follow.
Data Sharing for Bookings
The Department reported to the Office of Management and Budget, in documents supporting the FY 2005 budget request, that JABS eliminates the repetitive booking of offenders. We found, however, that the three booking evolutions at the arresting component, then the USMS, and then the BOP, continue to be necessary.
We observed demonstrations of the automated booking systems at participating agencies, including the interfaces between the legacy or missions support systems and the ABS, and interviewed participating agency officials and booking officers to obtain an understanding of how booking officers use the workstations. We determined whether information that existed in JABS could be used by the PTS and SENTRY. We also determined the number of data elements commonly associated with JABS bookings that are recorded more than once.
Users at the ATF, BOP, USMS, and DHS create their booking records by first recording booking information in their legacy systems: the ATF's N-FORCE, the BOP's SENTRY, the USMS's PTS, and the DHS's ENFORCE/IDENT, which are described in the section of this finding headed Participating Agency Automated Booking Systems. Booking officers then proceed to the agency's automated booking system, which contains JABS-specific functions and screens that have been integrated into the user interface at these agencies. Data elements needed for JABS that have been entered already into the legacy systems are automatically populated in the ABS. The booking officers then record in the ABS screens, fingerprints, photographs, and any additional data elements that are required for JABS. These procedures allow the ATF, BOP, USMS, and DHS to enter booking information once within the agency to create their booking records. However, for the USMS and BOP, the booking information may already exist in the central JABS database.
The DEA and FBI create their booking records by entering booking information only into their automated booking systems: the FBI's JABS and the DEA's FBS. Booking information recorded in the DEA's FBS is shared as part of the Firebird Case Management System. Booking information recorded in the FBI JABS is not shared with any other FBI applications.
Both agencies that perform subsequent booking actions, the BOP and the USMS, re-enter all of the information that has been recorded already by the ATF, DEA, FBI, or DHS, unless the USMS is the arresting agency, or the arresting agency has chosen not to book an offender at its own location before the transfer of custody to the USMS. (Some local facilities may be inadequately secure for prisoners, or inappropriate for other reasons.) The interfaces between legacy systems and the ABS were built in one direction only, from the legacy system to the ABS.
To measure roughly how much redundancy existed in booking data entry, we compared specific data elements recorded in the FBI's JABS, data elements recorded manually on hard-copy forms required by the USMS in one location, and data elements entered into the USMS's PTS. We counted as one action the manual recording of basic information (such as the offender's name), which may be recorded more than once on various hard-copy forms. This analysis indicated that 59 data elements had been recorded at least twice by the arresting agency and the USMS by the time the USMS had completed its booking process, and 49 data elements had been recorded three times. The 59 elements included basic identifying information about the offender, including names, address, physical description (height, weight, hair and eye color, scars, marks, tattoos), employer information, arrest information, and medical conditions and medications.
We did not include the BOP in this analysis because of its delay in implementing JABS, but the BOP will likely add a fourth time each of these elements is recorded. We also did not include fingerprint and photograph data in this analysis, because each agency verifies the identity of the individual through IAFIS.
Most users have direct access through their ABS to JABS records for bookings that have been performed only by their own component, or a subset of those bookings. The FBI and USMS have ABS access to JABS records that have been created only at each individual workstation. Users do not have access through their automated booking station to JABS records that were created by other components.
Access to JABS records that were created by other components is possible only through the JABS Query Tool, which is accessed independently of the ABS. The Query Tool is for inquiry only, and cannot be used to change, correct, delete, or otherwise manipulate data in the JABS database. Authorized users can view and print data from the Query Tool, except for fingerprints. In the future, the Query Tool is supposed to allow users to download data, import it into their ABS, and use it to create booking records, which could reduce data entry during second or third-tier bookings (at the BOP and the USMS). However, because ABS booking records for the BOP and the USMS are initiated in the components' legacy systems, there is little motivation for them to download existing JABS data into the ABS.
We asked why the interface between the USMS's ABS and the PTS was initially built in a way that required USMS booking officers to re-enter the information, rather than being designed to migrate data from the USMS ABS into the PTS. The JABS Program Manager explained that the USMS used an outdated database management system and wanted to update its architecture before building an interface.
The Program Management Office staff told us that they are beginning talks with the USMS about an interface to populate the PTS with data already in JABS, and the BOP staff told us they may also be interested in doing this. The Program Management Office staff also told us that the newest, web-based version of Core JABS (3.0) should support this shift and simplify access to the Core JABS database from the booking stations.
Data Sharing for Investigations - The JABS Query Tool
The Core JABS maintains data submitted by participating agencies in a repository that can be accessed by authorized users nationwide. Sharing booking data is intended, in part, to allow investigators to locate information about offenders quickly. The Core JABS Functional Requirements Document, which defines what the Core JABS is supposed to do, specifies several functions related to this responsibility, including to:
The JABS program provides authorized users, including investigative personnel, access to the nationwide database through the JABS Query Tool. Authorized users in a few agencies that do not perform bookings and do not need an ABS can access the JABS database through the Query Tool for investigative and research purposes.
We observed demonstrations of the Query Tool at the JABS Program Management Office and participating agencies, interviewed participating agency officials and booking officers, and reviewed the Query Tool User Guide to obtain an understanding of how it functions. We also used the Query Tool to obtain sample records for use in verifying that specified data elements exist and are being populated, and in a comparison of data values between the JABS database and the PTS.
We found that the Query Tool allows users to search the nationwide JABS database for persons based on identifying information, physical characteristics, and dates, agencies, and locations of bookings. Authorized users can view and print records from the JABS database, although they cannot currently view or print fingerprints. The FBI told us that being able to view and print fingerprints are functions it wants because many agents know how to compare fingerprints, and this function would be useful for investigations. The Program Management Office indicated the function had been turned off because there had been problems printing to various types of printers, some of which were not certified by the FBI for printing fingerprint cards.
The Query Tool also does not yet generate electronic photo line-ups for use by investigators from the Core JABS database. We were told by users in arresting agencies that this would be extremely helpful to them. It is possible that the Query Tool would find more widespread use were this made available to investigators.
The specific objectives and expected results related to the offender tracking system have not yet been fully defined by the JABS program. The Boundary Document indicates JABS will enable users to track offenders through disposition, but does not include performance measures for this goal. The Program Management Office told us that it had not focused on establishing the offender tracking system yet because the priority after September 11, 2001, had been to deploy the booking stations. The Acting Program Manager told us that no work has been scheduled yet regarding the offender tracking system but that work will be completed at some future, but as yet unspecified, time.
The Core JABS Functional Requirements Document specifies some requirements related to tracking, including requirements to store:
We believe an effective tracking system for federal offenders requires a complete and accurate database, meaning that all records should be included that meet some definition of completeness, so users know what offender records they can expect to find in the database. Information found in the records should also be accurate enough to be useful to law enforcement users who need to identify an offender's status and location until release. For critical information to be accurate, it needs to be current.
We found that the program had begun to establish the tracking system by creating the database and defining the requirements above, but had not defined exactly what records will be found in the JABS database. Additionally, not all of the data elements needed for tracking had been implemented, such as the disposition of the case, the date a sentence is due to expire, and a history of changes to an offender's location. Critical information may be outdated, because the JABS repository receives updates only from users who choose to send them, and none from the BOP. The details supporting this conclusion follow.
Records Included in the Database
Not all Departmental bookings are currently being recorded in JABS. The program has not defined what will constitute a complete JABS database, and has not planned how to achieve any specific objective in terms of what records are to be included. However, the percentages of bookings being captured by JABS are increasing for most Department components as deployment expands, as discussed in Finding 2, but have not reached the 100-percent mark.
The Department itself has not required that bookings be recorded in JABS, and not all Department components require this, even where the system is deployed. The ATF intends that all bookings be entered eventually, and DEA users have been instructed to book everyone through the FBS where it is deployed. FBI officials indicated that field offices had not yet been formally required to record bookings in JABS because not all offices were equipped to comply. The USMS and BOP record all booking transactions in their legacy systems, which should result in a JABS transaction being submitted at deployed locations. Most agencies use JABS voluntarily.
During interviews with representatives of the JABS Board of Directors, we were told that none of the components that make arrests, with the exception of the FBI, had plans to deploy JABS stations to 100 percent of their facilities. Even though the JABS budget pays for the purchase and installation of ABS equipment at Department components, the USMS representative told us that acquiring funds to pay for installation of the necessary high-speed communication lines was difficult, as other initiatives had higher priority. An ATF representative said the budget would run out of installation funds before it could install stations to all the sites the ATF would like.
Prior to implementation of the USMS's interagency booking capability, the arrests recorded in IAFIS through JABS were limited to arrests recorded at locations with an ABS. None of the Department's law enforcement components that make arrests had a policy requiring that offenders who were arrested at sites without JABS stations be entered into the system. ATF staff told us that, at some point in the future, agents who make an arrest but do not have access to an ABS may be required to send hard-copy documents to the nearest ATF facility equipped with an ABS so that booking information can be entered into JABS.
With implementation of the USMS's interagency booking transaction, it should be unnecessary to install an ABS at locations where other Department components book offenders to ensure that all Department bookings are recorded in JABS. Instead, an ABS with interagency booking capability could be deployed to all USMS detention facilities. This would capture nearly all offenders who have been detained or arrested by Department law enforcement officers because virtually all offenders who are charged with federal offenses are brought to and booked at USMS detention facilities soon after their arrest.
In addition to the fact that not all current arrests are being captured in the JABS repository, the program managers initially told us that they had no plans to populate the nationwide database with historical data on any population. However, in March 2005 program officials told us that the BOP has begun entering data on the current prison population. The usefulness of JABS in tracking the status and disposition of offenders will be limited until complete prisoner data is entered.
Data Elements for Tracking
For those records that are already in the JABS repository, we determined whether the data elements planned for inclusion in the repository had been implemented as specified in the JABS Interface Control Document.18 We included data elements that would be needed to track an offender through the system. The first question we considered was whether the data element itself is reflected in Query Tool records.
We compared the list of data elements from the Interface Control Document with a sample Query Tool record. We found that not all the data elements needed for tracking and specified in the Interface Control Document are reflected in the Query Tool record. These elements include court offense, court disposition, and disposition date. Also not found on the Query Tool record were issuance and expiration dates of identifying documents, alias names on identifying documents, vehicle identification numbers, and employer information.
The second question we considered was whether existing data elements were being populated. We selected 10 cases from the JABS database for each of the major components (ATF, BOP, DEA, FBI, USMS, and DHS) and documented the data elements being populated by each. The data elements for sentences and sentence expiration dates had not been populated for any offender by any agency. We found that additional data elements had not been populated by at least some participating agencies, including offender residence phone numbers, marital status, ethnic code, educational level, arms associated with the offender, fingerprint dates, and location of arrest. Some fields in the Query Tool record do not have corresponding fields in the ABS applications to enter data. For example, the FBI ABS does not offer a data element to record offender phone numbers, ethnicity code, or marital status.
There are three specific requirements identified in the Core JABS Functional Requirements Document needed to track offenders for the long term that have not been fully implemented. The requirement to store the location where the offender is serving a sentence and the date of arrival at the facility has been implemented only for the initial commitment to the BOP. The requirement to store the disposition of each case and the date that the sentence is due to expire has not been implemented. The requirement to store a history of changes to an offender's location has also not been implemented. If the offender is moved, the location stored in the JABS database will not be current.
Accuracy of Information
Updates and changes made in legacy systems, including the PTS and SENTRY, that affect data values in JABS are not automatically sent to JABS. Updates to records in JABS are sent when users choose to send them.19 We interviewed users and reviewed user manuals to determine if users are required to submit updates, and determined that there are no such requirements.
Updates and changes are especially relevant for the USMS and BOP, which record updated information regarding court charges, case dispositions, sentences, and sentence expiration dates in the PTS and SENTRY. Not updating JABS when changes are recorded in these legacy systems will cause JABS data to become outdated over time. A recent enhancement to the PTS/USMS ABS interface provides for the migration of updated data to the ABS, when a user pulls up a record that has been changed in the PTS. But pulling up the record in the ABS and sending the JABS transaction to the Core JABS depends on the user taking action. No changes entered into the SENTRY system are being sent to JABS, according to BOP officials.
We compared information in 30 records between the PTS and JABS, using the Query Tool. The differences we identified in data elements that appeared to be migrated initially from the PTS, and therefore likely to be the consequence of changes in one system and not the other, included the following:
We also found case disposition information in PTS records that was not shown in the Query Tool record.20
During our audit, the JABS program partially met its objectives for Department law enforcement users. The JABS program had: 1) automated the booking process for Department components, thereby ensuring the rapid and positive identification of offenders at deployed locations, 2) enabled users to share information by viewing and printing information created by other components, and 3) begun establishing the offender tracking system.
JABS automated the collection of offender data and provided an automated interface with IAFIS at deployed locations. JABS reduced the time needed to submit information to IAFIS, and the IAFIS response time has been cut back from weeks to hours, or even less time, at deployed locations. JABS had also reduced some redundant steps within agencies. Each booking station allowed users to print multiple fingerprint cards after a person's prints were successfully scanned once into the ABS, reducing the number of times a person has to be fingerprinted.
JABS, however, has not yet resulted in data sharing capabilities that would allow components to eliminate redundant steps, data entry, or data between components. Both the USMS and the BOP re-enter data on offenders that has already been booked in JABS by other participating agencies because they cannot import data from the ABS into the PTS and SENTRY.
The JABS Query Tool facilitates data sharing by allowing users to view and print JABS data, except for fingerprints. It was supposed to allow users to create electronic mug shot line-ups and view and print fingerprints; however, these functions were not available during our audit.
The offender tracking system has not been fully planned and the Core JABS data repository does not reflect all data needed for tracking offenders through disposition. Not all current arrests by the Department are being recorded through JABS, so the information is incomplete, and no goals have been established defining who should be included in the database. Additionally, corrections or updates to data contained in the repository are not necessarily updated in JABS, even when they are made to component legacy systems. This will have the effect of JABS data becoming outdated over time.
We recommend the Justice Management Division:
The performance measure related to deploying JABS as stated in the JABS Boundary Document is to "deliver 100-percent connectivity to components' automated booking systems." Program officials told us that this performance measure is interpreted to mean that JABS is capable of transmitting and processing bookings with each component's automated booking system, even if only one station is deployed in a component, which has been achieved and is described in Finding 1.
There are no objectives establishing that all persons, or which specific persons, detained or arrested by the Department's law enforcement components should be booked through JABS. There are also no formal goals establishing the number or type of sites at which the booking stations should be deployed. Program officials told us that there is no goal for the program to make the system available for all Department bookings.
JABS program officials explained that their role was to respond to the needs of Department components and other participating agencies by providing JABS booking stations and services based on the components' requests. Program officials also told us that the components identify locations to be equipped with JABS stations and determine who should be booked through JABS. Booking officers are expected to use it voluntarily in most components. The specific objectives for deployment, therefore, are established by the components and other participating agencies when they identify sites to be deployed.
JABS program officials told us that priorities for the program shifted after the terrorist attacks of September 11, 2001, when only the DEA's automated booking capability had been linked to JABS.21 Resources were diverted away from reducing duplication and making other improvements to JABS to implementing the system at additional components and deploying stations. Other aspects of the JABS mission were postponed until all components were linked to JABS and the number of stations requested could be fulfilled.
Nationwide and Core JABS
As of November 2004, the JABS program had provided automated booking capabilities to all of the Department's law enforcement components, and to about 77 percent of the components' requested locations, or 841 Department locations, out of a total 1,090 sites planned for inclusion in the system.22 The implementation began with an early version of the system at the DEA in July 2000, followed by the USMS and FBI in August 2002, and the ATF, which started a pilot office in December 2003. The last Department component to begin JABS operations was the BOP in April 2004, but all of its locations had been equipped with JABS stations in the summer of 2002. Additional details about implementation are found in Appendix III. All locations that have been requested by the DEA as of November 2004 have been linked to JABS. The deployments for the other components based on their requested sites have not been completed.
JABS was also implemented at the Immigration and Naturalization Service (INS) beginning in September 2002, as part of its efforts to make FBI fingerprint data available to immigration and border officials. The JABS Program Management Office provided documentation showing that the system had been deployed to 121 locations by the time the INS became part of the Department of Homeland Security (DHS) in March 2003.
In addition to deploying automated booking systems to law enforcement components, the JABS program had to develop a data repository and conduit to IAFIS in a way that would allow integration with the components' mission critical systems. These central functions of the Core JABS had to be operational before any component's automated booking system would be fully operational. The Core JABS first became operational in July 2000, with implementation and linking of the early DEA booking system.
Deploying to Department Components
The Program Management Office has completed interfacing the automated booking systems with JABS for the five Department law enforcement components, (the ATF, DEA, FBI, USMS, and BOP), and the DHS. To assess the extent of deployment, we analyzed information and data provided by the JABS Program Management Office and components, interviewed component and program management staff members, and conducted visits to the ATF, BOP, DEA, FBI, USMS, and DHS. We determined the number of sites and stations requested by Department components and non-components, and confirmed deployment at each site using transaction reports.
We met with the JABS Program Management Office to determine the number of sites each component had identified to be equipped with JABS stations. To verify the figures provided by program officials, we independently interviewed members from each of the Department's law enforcement components. We also contacted component staff to discuss the level of deployment and, if applicable, the reasons for not deploying JABS stations to all available sites. We found the number of sites requested by components to be consistent between the Program Management Office and components.
The general sequence of events for deploying booking stations is for the Program Management Office to receive a request for booking stations from a Department component or other participating agency. Except for the DHS, the program acquires, physically installs, and tests each booking station, and connects it to the Core JABS so transactions can be processed. After booking stations are installed and connected, the program provides training to the component, if requested. Some participating organizations conduct their own training.
The chart below compares the number of sites requested by each component with the number of sites deployed, as of November 2004. Deployed means not only that the workstations have been installed with all elements functioning, but also that the connections with the Core JABS have been implemented and the site has the ability to submit transactions.
Comparison of the Number of Deployed Sites to the
As reflected in the table above, the degree of deployment completed varies significantly by component. The DEA, at 100 percent, and the ATF, at 25 percent, reflect their roles as the earliest and most recent installation starts. Since the number of sites requested changes over time, the statuses on the table may change frequently. Specific deployment information for each Department law enforcement component follows.
ATF - The ATF began using one workstation in January 2004, at Richmond, Virginia, as part of a pilot test, but then delayed further implementation of JABS until a new version of the system was developed and implemented. This new web-based version, JABS Version 3.0, Three-Tier Architecture Plan, received certification and accreditation to operate in June 2004. By spring 2004, the ATF had requested that workstations be installed in an estimated 146 offices, and by November 2004, the ATF had 37 sites connected to JABS. The JABS program and the ATF have projected that between 130 and 140 total ATF sites will be deployed by September 2005, which will satisfy the current request.
ATF officials we interviewed said the ATF had 389 offices. The ATF selected JABS sites from offices that perform at least about 50 bookings per year. When asked why a 100-percent implementation was not requested, the ATF representative stated the budget did not support that, and it would be too expensive to install equipment at all locations. It is also the ATF's expectation that every person arrested would be booked either at an ATF site equipped with JABS, or the offender would be taken to the closest USMS site for booking. If a USMS facility is not feasible, paperwork for the offender would be sent to the nearest ATF facility for input.
BOP - Between June and August 2002 the program installed 240 JABS workstations at 139 facilities.24 The system should have been available for use soon thereafter, but it failed when placed into production at six locations. The Program Management Office determined it would be necessary to scrap, redesign, and rewrite the software that had been delivered by the contractor. Additional discussion of this problem is in the section of this finding headed Delayed Deployment at the BOP. The first transactions from the BOP following implementation of the new system were performed in April 2004, almost two years later.
The JABS program and the BOP agree that the component is fully installed, with no additional sites planned. As of July 31, 2004, transactions had been submitted to the Core JABS from 87 of 139 sites, and by November 2004 the BOP had 120 of the sites reporting transaction activity. When the BOP begins operations in the remaining 19 sites, it will be fully deployed according to current projections.
DEA - The DEA was the first Department component to automate the booking process and connect with the Core JABS in July 2000. According to a DEA official, the early implementation proceeded slowly until after the terrorist attacks on September 11, 2001, when emphasis was placed on rapid deployment. By the end of FY 2003, 202 of the DEA's 285 requested sites had been deployed, and by November 2004 all 285 were deployed. The JABS program and the DEA consider domestic deployment for the DEA to be complete, although there are about 313 total DEA sites. (See the table on page 43.)
At the end of July 2004, the DEA's ABS had been deployed in 9 foreign locations, including Brazil, Greece, Mexico, and Pakistan. Officials of the JABS program and the DEA told us the DEA is interested in installing about 75 additional stations internationally, but no information has yet been made available to confirm their request or status.
FBI - An FBI official told us the agency had initially decided to make JABS stations available widely throughout the agency, so originally the FBI requested JABS stations for all of its 448 field office and resident agency sites. By the end of November 2004, the program had deployed stations to 235 FBI sites. An official of the FBI's Criminal Justice Information Services (CJIS) Division said the FBI intended for all of its offices to use the workstations for prisoners, confidential informants, or both.25 Since the FBI had not deployed JABS to all locations, it had not yet imposed requirements about booking prisoners or making inquiries about confidential informants through JABS.
The FBI official told us the agency was considering a policy requiring FBI agents to use the booking stations even if they do not bring prisoners to an FBI location. Agents will enter data, or download it from the Query Tool, and scan in hard-copy fingerprint cards printed by the USMS to submit the booking to IAFIS.
As of November 2004, the FBI was reassessing its original request for 448 workstations and the need for JABS equipment at specific locations, because of mobile JABS booking stations and the new interagency booking function now available at USMS locations. Mobile workstations will allow the agency to deploy stations where they will be the most useful for limited periods. The USMS's new interagency booking function may make it unnecessary for every FBI office to have JABS equipment because a booking can be recorded and sent to IAFIS as an FBI booking without having to be sent directly from an FBI office. In March 2005 the Acting Program Manager told us that the FBI had determined that 235 sites would meet its needs.
The FBI had recently received a number of mobile JABS workstations. As of November 2004, the FBI had placed 18 mobile stations at 4 sites: 1) Clarksburg, West Virginia; 2) Quantico, Virginia; 3) New York City; and 4) Rolling Meadows, Illinois. These can be used at some locations in lieu of regular workstations, allowing the FBI to deploy stations where they will be the most useful for limited periods. An FBI official said they might decide to shift some resources to international sites. By the end of July 2004 the FBI had booking stations deployed in 4 foreign locations, including Afghanistan, Iraq, and Turkey.
USMS - JABS became available to the USMS beginning in August 2002. The USMS had requested that JABS stations be installed at 285 facilities. By the end of November 2004, the JABS program had deployed workstations to 164 of the sites. The USMS representative to the JABS Board of Directors confirmed the program's plans that deployment to all of the requested 285 sites will be completed by the end of FY 2005. The 285 sites, however, do not represent all USMS facilities at which bookings occur. The USMS representative said the USMS envisions that every office with a staff of 5 or more people would eventually receive a JABS station.
Deployment plans for the USMS depend on upgrades to the component's network and communications capabilities. The USMS network is old and high-speed communication lines (T-1 lines) needed to fully deploy the system have not been installed at many locations. The USMS representative told us that it has had problems obtaining funding to upgrade its network and that only about half of approximately 336 total USMS offices had been upgraded with high-speed lines as of summer 2004. Additionally, the USMS representative told us that smaller sites will probably not get upgraded lines at any time in the near future.
In addition to the impact on USMS operations, this resource constraint has prevented the JABS program from maximizing the USMS's role as the central processor of federal offenders. Since interagency booking functionality was implemented through the USMS in June 2004, any federal law enforcement agency can submit its bookings through the USMS automated booking system to JABS and IAFIS, and receive the IAFIS response electronically. If JABS were deployed at every USMS location to which offenders are brought by arresting agencies, JABS and IAFIS would be available electronically to any federal law enforcement agency with access to a USMS site. The USMS and the JABS program have informally discussed the possibility of using JABS funding to help implement the communications upgrades, but no decision had been reached at the time of our audit.
Other Participating Agencies
The JABS program will provide services to any federal agency that is required to submit fingerprints to the FBI. The only organization outside the Department that uses JABS for a significant number of bookings is the DHS. The National Institutes of Health Police and two offices of the U.S. Army Police have also connected to JABS to submit bookings. Several additional agencies have expressed interest, and JABS program officials are in discussions with the agencies identified below. A handful of organizations are connected to the JABS Query Tool for purposes of accessing information only, but not to perform bookings.
DHS - The JABS program deployed 198 JABS stations to 121 INS sites while the INS was still part of the Department. When the INS became part of the DHS in March 2003, responsibility for deployment transferred to the DHS as part of the IDENT/IAFIS project, which is now part of the US-VISIT Program Office at the DHS. The DHS officials we interviewed said that the only service the JABS program provides DHS is the secure conduit to IAFIS from their booking stations. DHS officials told us their goal is to deploy its automated booking stations to all Ports of Entry and Border Patrol stations as part of its US-VISIT initiative. The DHS anticipates that all bookings will be processed through JABS, which it calls IDENT/IAFIS.
Others - As of March 2004, JABS workstations were installed and linked to the Core JABS at two locations of the U.S. Army Military Police and at the National Institutes of Health Police for purposes of booking offenders. At the time of our audit, plans were in place to link the U.S. Air Force Office of Special Investigations, and discussion had begun with the Secret Service to implement JABS. As of mid-December 2004, neither organization had connected to the system. An additional link with the U.S. Courts is anticipated in December 2005. Other federal agencies have read-only authorization to use the nationwide JABS database, which they access through secure accounts using a web browser.
Discussions between the JABS program and other organizations outside the Department have occurred, with agreements to deploy stations in some cases. The organizations listed below have expressed an interest in obtaining JABS stations.
We analyzed information about funding, budgets, expenditures, and obligations for implementation of the nationwide JABS from FY 1999 through the first quarter of FY 2004. The figures do not include funding for the JABS Pilot Project that was terminated in July 1999.
The JABS program received $82.68 million in funding for implementation and operations through FY 2004 from the Asset Forfeiture Fund (AFF), congressional appropriations, and the Department's Working Capital Fund (WC). The table below reflects the funding by fiscal year. The program allocation in the current version of the FY 2005 budget is $20.185 million.
The JABS program has reimbursable agreements with the DHS and U.S. Air Force for workstation deployments and operational services provided by the Core JABS. The funding figures above do not include the amounts that are paid to the program from these agreements.
The JABS program has obligated, from FY 1999 through the first quarter of FY 2004, $72,879,776 for nationwide JABS implementation, as indicated in the following table. The $72.88 million total amount includes costs that will be reimbursed to the program through the reimbursable agreements with the DHS and U.S. Air Force.
For the tables above and below, the "Program Management Office Operations" category includes, but is not limited to: 1) office supplies, 2) salaries and benefits, 3) travel, and, 4) office equipment. The "Core JABS" category includes design and engineering effort, security, and operations and maintenance. The "Amounts for Component Support" category represents component booking station equipment and work coordinated by the Program Management Office on the components' automated booking systems.
Actual expenditures are shown in the table below. The JABS Program Management Office had expended, for fiscal years 1999 through the first quarter of fiscal year 2004, $57,459,127 for nationwide JABS implementation. The $57.46 million amount includes the reimbursable agreements with the DHS and U.S. Air Force.
The original plan for JABS called for each component to build its own automated booking system. The Program Management Office became responsible for the total JABS budget early in the project, and has coordinated the development and deployment of each component's automated booking system.
The participating agencies' support amounts are broken down further by agency. Each participating agency's support obligation and expenditure amounts for the period FY 1999 through the first quarter of FY 2004 are reflected in the next table.
Participating Agencies' Assistance Obligations and Expenditures
We reviewed transaction reports generated from the JABS database to determine the types and volumes of transactions that are being processed by the Core JABS.27 The following table summarizes transactions showing bookings, quick searches, and all transactions. The number of JABS transactions increased from 6,015 in FY 2002 to 711,141 through July 2004. The number of transactions is projected to increase to at least 3 million in FY 2005, and 6 million in FY 2006, as new users are brought into the system.
Total JABS Transactions
The majority of transactions since FY 2002 have been the quick search/TPRS requests submitted by the DHS. Bookings through three quarters of FY 2004 accounted for about 18 percent of all transactions, with the DHS quick searches making up about 72 percent. Other transactions, such as inquiries and updates accounted for the remaining 10 percent of transactions.
Percentage of Component Arrests Booked in JABS
The JABS Program Management Office does not determine or track the percentages of Department arrests being captured as bookings through JABS. To determine what percentage of all arrests were being recorded as bookings in JABS, we requested arrest data from the DEA, FBI, USMS and DHS. We did not request arrest data from the ATF, as only the pilot office operated a JABS station during our audit period. We also did not request commitment data from the BOP because the BOP had just begun using the system in the spring of 2004. As of December 2004, the DHS had not provided the arrest data we requested. We compared the arrests reported by the DEA, FBI, and USMS to each component's bookings for FYs 2002 and 2003, and through the 3rd quarter of FY 2004. The bookings performed through JABS as a percentage of each component's reported arrests are shown in the table below. For example, for FY 2003, about 24 percent of the FBI's arrests were booked through JABS.
Percentage of Component Arrests Booked in JABS
It is clear that many arrests made by the Department were not being recorded in JABS as bookings through July 2004, and that the wide deployment has not ensured that all Department arrests are being booked through JABS. We found that 40,748 of 60,129 arrests (68 percent) by the DEA, FBI, and USMS were recorded in JABS as bookings for the first three quarters of FY 2004. The significance of an arrest being submitted to JABS as a booking is that types of transactions other than bookings are not stored in IAFIS. The USMS would have submitted criminal history transactions to JABS for many of the arrests the components had not recorded in JABS. Criminal history transactions, however, are not stored in IAFIS, so do not result in the quick addition of offender records in IAFIS. Until implementation of the USMS's interagency booking capability, the USMS was not able to submit a booking on behalf of other components to be stored in IAFIS.
With the possible exception of the USMS, however, we found that the percentage of all arrests being reported as bookings through JABS is increasing. The substantial increase in the percentage of bookings being completed through JABS is a direct result of the continued deployment since FY 2002, when only 103 sites were reporting transactions nationwide. The USMS figures indicate a decrease in FY 2004 from the prior year, but USMS representatives told us that they believe this is the result of reporting changes that were implemented in FY 2004, and that had the effect of reducing the reported number of bookings based on USMS arrests.28
As of September 2004, there were 24,988 potential users, for all organizations, with active passwords. The total number of potential users, organization or component, and their percentage of the whole, is shown in the following graph.29
Among the major law enforcement users of JABS, the DHS has the largest number and percentage of authorized users (56 percent). Not all authorized users have used the system because they may have no need to book anyone for long periods. DHS immigration inspectors and border patrol officers have to be authorized to access the system as they are subject to being assigned to work in secondary inspections, where JABS is used. We expect the number of ATF users to increase dramatically as deployment continues. Most of the 187 "Other" users shown in the chart above have access to the JABS database through the Query Tool for investigation and research purposes, not for booking offenders.
We identified a nearly two-year delay in implementing the system at the BOP after all the requested equipment had been installed. All of the 240 workstations that were requested by the BOP were installed between June and August 2002. However, the BOP only began submitting JABS transactions successfully in the spring of 2004, nearly two years after installation. In addition to losing the use of those workstations and the additional records that would have been created in the JABS database, BOP ABS software had to be redesigned, and the JABS program paid the contractor for two design and development efforts.
JABS program and BOP officials told us that the software resulting from the initial Design and Development (D&D) work done on behalf of the BOP had major problems that were not discovered until after the program had installed all 240 workstations at the BOP's requested sites. The Acting Program Manager told us that the problem occurred for two reasons: 1) the Program Management Office did not perform oversight to a sufficiently detailed level to identify this problem during the software development stage, and 2) the contractor did not follow the processes described in its quality assurance plan.
Program officials told us that a contractor assured them the design and development effort was proceeding appropriately. A program official also said the initial system had passed system acceptance testing, and 6 sites had been connected when the failure occurred. However, the system failed when more than a few users tried to enter data. The BOP and JABS officials told us they tried to correct the problems, but decided it would be more cost effective to scrap the existing code and start from scratch due to its inappropriate design.
The JABS Program Management Office was not able to provide documentation to confirm that it had effectively monitored the D&D efforts being performed by the vendor. The Acting Program Manager told us that the Program Management Office participated in design reviews but relied on the contractor to review the programming code in accordance with the contractor's quality assurance plan.
BOP Early Installation Costs
We calculated the cost of the early installation of the BOP's JABS equipment using financial documents provided by the JABS Program Management Office.
The cost of the equipment was about $4.2 million. We calculated this by adding the following amounts shown in the JABS FY 2002 Spending Plan:
In addition to the purchase cost of equipment, records show the Program Management Office spent about $176,000 for BOP travel and training for the deployment of the stations.
The initial D&D effort to integrate the BOP with JABS was contracted out. The contractor provided a senior information systems engineer who worked on the initial project a cumulative 859 hours, which concluded in September 2002 at a total staff cost of $88,082.
The second BOP D&D effort was also contracted out. The contractor concluded work on or about January 31, 2003. A second contracted senior information systems engineer, with the assistance of four contracted software engineers, worked on the second integration effort for a cumulative 1,432 hours at a total additional staff cost of $131,457.
We determined the total costs associated with the early deployment and installation of JABS to the BOP were $5,042,933, as of February 2004. We also found that $88,082 was wasted by the contractor in the initial D&D effort because the initial work had to be corrected. We did not question this amount because it represented an immaterial portion of the total cost of the project.
BOP Inventory Confirmation
As discussed above, about $4.2 million was spent on equipment and supplies to deploy the 240 stations. As substantial assets were expended for the acquisition and installation of the equipment that sat unused for a long time, we asked the BOP to account for the installed equipment.
BOP booking stations had been installed at 139 sites during 2002. We reviewed a transaction report for May 2004 and determined that the BOP had submitted at least one transaction from 23 of the 139 sites at which the stations were installed. We contacted 115 sites that had not submitted transactions to confirm and account for JABS equipment, and to obtain serial/inventory numbers for their equipment. We compared these with corresponding identification numbers from a Shipping Status Report for the BOP equipment that was provided by the Program Management Office.
We received responses from all 115 BOP sites. The results of responses from the property confirmation letters were that:
Our analysis revealed only minor discrepancies, with unaccounted equipment valued at $2,019. We determined these discrepancies are immaterial and that the BOP has adequately accounted for installed equipment.
While conducting field work to confirm that the workstations being reported as deployed had, in fact, submitted transactions, we matched reported transactions for each site to the listing of deployed locations provided by the program. We noted that some sites had submitted either no transactions, or very few transactions, since they were deployed. We decided to determine whether each site actually had a JABS station, or if the limited activity was the result of problems the sites were having with JABS.
We identified 32 sites showing no activity: 10 DEA, 13 FBI, 4 USMS, 4 DHS, and 1 U.S. Parole Office.
The 10 DEA facilities that had no transaction activity were located at:
In June and July 2004, we contacted staff at each of the affected sites and inquired about the lack of activity. The results were that:
We also identified 13 FBI sites that did not record any activity. We used the same method to determine why a deployed station was not being used. The FBI respondents provided the following reasons for no activity.
To determine reasons some sites had submitted only a few transactions, we categorized sites as having low transaction levels by arbitrarily selecting a threshold of 15 transactions, based on cumulative activity as of July 2004. We considered any site with 15 or fewer transactions to have low levels of activity. We did not include the BOP in this review, as the agency was essentially not operational on JABS at the time of our review. We identified 73 sites with low transaction activity. Of the 73, 51 sites belonged to the FBI and 12 were DEA facilities. The remaining 10 sites were distributed between the DHS, the U.S. Army site at the Picatinny Arsenal, and one USMS facility in Victoria, Texas.
Of the questionnaires sent to DEA locations, 7 responded. Four sites attributed their low usage to the fact that they were small offices and consequently there were a minimal number of arrests. Staff from 1 site reported the USMS books their arrestees, thus reducing their transaction activity numbers. Four sites reported that inadequate training and considerable hardware problems had prevented them from using the station effectively.
As with the DEA, we mailed questionnaires requesting an explanation of the low level of activity to the 51 FBI offices. We received 32 responses. Staff from 20 sites attributed their low level usage to the low number of arrests they made due to the types of investigations they conducted, or to the fact that the office was small and had a limited number of trained agents. Nine respondents reported they take their arrestees to the USMS for booking. Five reported the lack of training as the primary reason for low usage in their office. Three reported equipment problems, and one reported they use the equipment only for photos and taking fingerprints for comparison. The sum of the respondents exceeds the 32 responses, but this is because several respondents provided more than one explanation in their answer.
All of this suggests that some of the workstations that have been deployed will not be used that much or at all, and that the program should focus on providing access to JABS at the USMS detention sites, where nearly all offenders arrested by federal law enforcement agencies are brought to be processed.
JABS Program Plans
The JABS program plans for expansion through FY 2006 include deploying additional workstations to the ATF, DEA, and USMS, and linking with the U.S. Air Force Office of Special Investigations, the U.S. Courts, and the Secret Service. The total deployments projected for the ATF and FBI through FY 2006 do not fulfill the total stations requested at this time, although the FBI is reconsidering its request for a booking station at each field location based on recent developments in the interagency booking function at the USMS and mobile booking stations. The program also plans to convert the USMS automated booking system to the web-based version of JABS by September 2005. The total projected costs for FY 2005 were $20.3 million to cover all planned activity.
Neither the components nor the JABS program could provide documentation to support the number of sites that would represent an optimal number of booking stations that would ensure that all or most bookings performed by the Department are submitted electronically to IAFIS through JABS. We attempted to determine an approximate universe of possible sites in the Department's law enforcement organizations that could use a JABS station, assuming there were no funding, logistical, or operational constraints.
During interviews, we asked component personnel for the number of offices in which it would be possible to use JABS. The second column of the table below identifies the number of sites that could be recipients of a JABS station. Our intent was to identify an approximate universe of sites should the Department be 100-percent equipped with JABS.
With the exception of the ATF and FBI, the number of sites that have been requested by the components is within range of meeting what would be full deployment of JABS for the Department, even if defined broadly to include JABS stations in virtually all offices where offenders are booked. Changes in JABS itself, however, may be decreasing the need or motivation for each location of each component to have its own JABS booking station.
Interagency Bookings at the USMS
Since interagency booking functionality was implemented through the USMS automated booking system in June 2004, any federal law enforcement agency can have arrests submitted through JABS to IAFIS, with the IAFIS response sent back to the arresting agency electronically. Before this function was implemented, a component had to submit an electronic booking from a workstation belonging to the component to be identified by IAFIS as the arresting agency and maintain responsibility for reporting case disposition to the FBI. This is because the USMS's automated booking system was previously not capable of assigning the arrest to another agency if the USMS submitted the first booking transaction in JABS.
We found that the interagency function was installed and operating at about 122 USMS facilities. The USMS can now submit booking packages that identify an agency other than itself as the arresting agency. The arresting agency receives credit for the arrest and retains responsibility for the disposition of the case. This feature allows other organizations to participate in JABS without having to build their own booking systems. The Program Management Office provided documentation indicating that, as of August 2004, the USMS had submitted 5,447 bookings on behalf of 34 different organizations. In addition to submitting bookings on behalf of the four Department law enforcement organizations that have JABS booking systems, the USMS also submitted bookings on behalf of the Department of Defense, Food and Drug Administration, Internal Revenue Service, National Park Service, Veterans Administration, and 25 other agencies.
If JABS and interagency booking capabilities were deployed to every USMS location to which offenders are brought by arresting agencies, JABS and IAFIS would be available electronically to any federal law enforcement agency with access to a USMS site, without needing to have their own JABS booking stations. This could ensure that virtually all offenders arrested by federal law enforcement agencies could be booked electronically through JABS and IAFIS. Deployment to all USMS sites, however, is not currently planned, and this potential has not yet been realized. The immediate problem, according to both JABS program and USMS officials, is funding for communications network upgrades. The USMS and the JABS Program Management Office have informally discussed the possibility of using JABS funding to help implement the communications upgrades, but no decision had been reached at the time of this review.
By the end of November 2004, the JABS program had successfully deployed automated booking capabilities to about 840 sites, or about 77 percent of the Department's law enforcement components' locations planned for inclusion in the system. The system, however, is not yet deployed or used in a way that ensures that all Department arrests are recorded in JABS or are submitted electronically to IAFIS.
The central point of contact for offenders in the federal booking process is the USMS. The JABS program has begun to take advantage of this by the recent implementation of interagency booking functionality. This capability can be used to maximize the number of federal bookings that can be electronically linked to the FBI's IAFIS and stored in the nationwide JABS repository for authorized users to view.
We also found that 240 JABS stations were installed at BOP facilities nearly two years before the system was usable because the contractor did not adequately review the software code and the Program Management Office's oversight was not sufficient to identify that the contractor was not following its quality assurance plan. The result of this problem was that the Department lost the use of the assets for almost 2 years and additional funds had to be spent redeveloping the software.
We recommend that the Justice Management Division: